The present invention relates to an apparatus and method for creating a music piece by interconnecting sound fragments.
Among the conventionally-known music piece creation techniques is a technique called “audio mosaicing”. According to the audio mosaicing technique, various music pieces are divided into sound fragments of short time lengths, so that sound fragment data indicative of waveforms of the individual sound fragments are collected to build a sound fragment database. Desired sound fragment data are selected from the sound fragment database, and then the selected sound fragment data are interconnected on the time axis to thereby edit or create a new music piece. Examples of literatures pertaining to this type of technique include:
[non-patent literature 1] Ari Lazier, Perry Cook, “MOSIEVIUS: FEATURE DRIVEN INTERACTIVE AUDIO MOSAICING”, [on line], Proc of the 6th Int. Conference on Digital Audio Effects (DAFx-03), London, UK, Sep. 8-11, 2003 [searched Mar. 2, 2007], Internet<URL: http://soundlab.cs.princeton.du/publications/mosievius_dafx—2003.pdf>; and
[non-patent literature 2] Bee Suan Ong, Emilia Gomez, SebastianStreich, “Automatic Extraction of Musical Structure Using Pitch Class Distribution Features”, [on line], Learning the Semantics of Audio Signals (LSAS) 2006, [searched on Mar. 6, 2007], Internet<URL: http://irgroup.cs.uni-magdeburg.de/lsas2006/proceedings/LSAS06—053—065.pdf>.
In order to obtain expressive music piece data, it is necessary to prepare in advance a variety of sound fragment data having various characteristics and select and interconnect suitable ones of the sound fragment data. However, finding desired sound fragment data from among the enormous quantity of the sound fragment data is very hard work.